Python - Splitting Criteria: Gini | Python Courses in Tamil | Skillfloor - video Dailymotion

Skillfloor

Learn the concept of Gini impurity, a key splitting criterion used in Decision Tree algorithms, with this easy-to-follow tutorial in Tamil!   Our Website: Visit 🔗 http://www.skillfloor.com  Our Blogs: Visit 🔗 https://skillfloor.com/blog/  DEVELOPMENT TRAINING IN CHENNAI https://skillfloor.com/development-training-in-chennai  DEVELOPMENT TRAINING IN COIMBATORE https://skillfloor.com/development-training-in-coimbatore  Our Development Courses: Certified Python Developer Visit 🔗https://skillfloor.com/certified-python-developer Certified Data BASE Developer Visit 🔗https://skillfloor.com/certified-data-base-developer Certified Android App Developer Visit 🔗https://skillfloor.com/certified-android-app-developer Certified IOS App Developer Visit 🔗https://skillfloor.com/certified-ios-app-developer Certified Flutter Developer Visit 🔗https://skillfloor.com/certified-flutter-developer Certified Full Stack Developer Visit 🔗https://skillfloor.com/certified-full-stack-developer Certified Front End Developer Visit 🔗https://skillfloor.com/certified-front-end-developer  Our Classroom Locations: Bangalore - https://maps.app.goo.gl/ZKTSJNCKTihQqfgx6 Chennai - https://maps.app.goo.gl/36gvPAnwqVWWoWD47 Coimbatore - https://maps.app.goo.gl/BvEpAWtdbDUuTf1G6 Hyderabad - https://maps.app.goo.gl/NyPwrN35b3EoUDHCA Ahmedabad - https://maps.app.goo.gl/uSizg8qngBMyLhC76 Pune - https://maps.app.goo.gl/JbGVtDgNQA7hpJYj9  Our Additional Course: Analytics Course https://skillfloor.com/analytics-courses https://skillfloor.com/analytics-training-in-bangalore Artificial Intelligence Course https://skillfloor.com/artificial-intelligence-courses https://skillfloor.com/artificial-intelligence-training-in-bangalore Data Science Course https://skillfloor.com/data-science-courses https://skillfloor.com/data-science-course-in-bangalore Digital Marketing https://skillfloor.com/digital-marketing-courses https://skillfloor.com/digital-marketing-courses-in-bangalore Ethical Hacking https://skillfloor.com/ethical-hacking-courses https://skillfloor.com/cyber-security-training-in-bangalore  #giniimpurity #decisiontree #pythontutorial #machinelearning #tamilcoding #skillfloor #datascience #pythonintamil #mlalgorithm #splittingcriteria #classification #datasciencetamil #machinelearningtamil #pythonprogramming #mlmodels #techeducation #codingintamil #featureselection #pythoncourse #tamiltech

Transcript

00:00Hello everyone. In this video, we will talk about splitting criteria Gini.

00:07We will talk about Gini in the decision tree and how to use Gini.

00:18In splitting criteria, first we will talk about Gini index.

00:22In the entire dataset, we will talk about a randomness or a disorder.

00:30We will talk about Gini index is 0, and we will talk about the proper data.

00:34We will talk about the change in the order, classification and classification.

00:38We will talk about randomness.

00:40If Gini index is high, we will talk about split.

00:45We will talk about decision tree and the spitting is proper.

00:50We will talk about decision final.

00:52We will talk about accuracy and provide accuracy.

00:55That's why we will talk about Gini index.

00:58We will talk about tree tree.

00:59The Gini index is basically what we call formula.

01:021 minus probability of each and every class labels.

01:06That is target column based.

01:08First, we will take a simple example.

01:11This is target column, whether it is an apple or not.

01:14We will decide yes or no.

01:16So, we will choose what we call color and size.

01:21So, if we call Gini index, we will talk about the entire data.

01:24Actually, we will talk about yes or no, mixed up.

01:27Then, we will talk about disorder column.

01:29We will talk about disorder column.

01:30We will talk about disorder column.

01:31We will talk about disorder column.

01:32Now, we will talk about probability of target column, yes or no.

01:37So, first probability of yes.

01:39How many data is?

01:403 data is.

01:41Out of 6 data.

01:42Probability is 3 by 6.

01:43That's why probability of no is 3 by 6.

01:46This is 0.5.

01:47So, if we calculate the Gini index index,

01:491 minus probability of yes whole square plus probability of no whole square.

01:54Okay?

01:55So, in the CI in red, we have target column.

01:57So, targets order class labels.

02:01Okay?

02:02So, we will substitute value.

02:06So, 1 minus 0.25 plus 0.25.

02:081 minus 0.5.

02:09We will find 0.5.

02:10We will find 0.5.

02:11So, we will calculate the Gini index index index.

02:14Okay?

02:15So, we will calculate the Gini index index index.

02:18Now, we will consider a data set.

02:20That means, we will paste the tree structure and paste the Gini index index index.

02:23So, first, we will consider the weekend data.

02:26Then, we will select the weather, parent, money column.

02:29So, we will paste the decision.

02:31We will provide a decision on the weekend.

02:33We will go to cinema, tennis, or stay in the street.

02:35We will go to shopping.

02:37So, now, what we will do is,

02:40we will create a tree structure.

02:41So, what we will do in the tree structure?

02:43We will create a root node.

02:44Then, we will create a decision node.

02:47Then, we will create a decision node.

02:48Then, based on the final,

02:49we will choose for that division.

02:51Correct?

02:52So, the first root node,

02:53we will select the root node.

02:54We will not have any type of root node.

02:56Like, weather, parents, money, etc.

02:58So, in the three different columns,

03:00we will calculate the Gini index index.

03:02If there is a minimum value,

03:04we will decide the root node.

03:06So, first, step 1.

03:08Then, we will select the Gini index index for this entire dataset.

03:10In the entire dataset,

03:11if we look at the six data,

03:12we will pick up the cinema and say,

03:14two data is tennis,

03:16one data is shopping,

03:17and one data is stay in.

03:19Then, if we calculate the Gini index index,

03:22the value is 0.58.

03:24Now, first, we will choose root node.

03:26So, we can choose root node. So, if we look at entire data results, we can choose 58% disorder.

03:32So, we can create root node based tree structure. So, first, we will consider weather column.

03:37So, in this weather column, what are the possibilities?

03:42Sunny, rainy and windy.

03:47There are three data. So, in sunny, most of the data, there are three data.

03:51So, then, windy, 4, and rainy, 3. So, first, sunny data, consider.

03:58So, sunny data, w1, w2, w3.

04:02So, here are three data, decision node.

04:06Cinema, tennis and tennis. So, here are two tennis and one cinema.

04:13So, sunny data, where are three instances?

04:16So, what is the name of the cinema? 1 by 3.

04:18Tendence is 2 by 3.

04:20So, we will check the guinea index for sunny.

04:23So, first, weather, what is the category?

04:26But, in one category, we will calculate the guinea split index.

04:29So, we consider sunny data.

04:311 minus 1 divided by 3 the whole square, plus 2 divided by 3 the whole square.

04:37We will calculate value.

04:38The value is 0.44.

04:40Okay?

04:41That means windy.

04:42In windy, there are three cinemas and one shopping.

04:45So, in the cinema, it is 3 by 4.

04:46Shopping is 1 by 4.

04:47So, this is the guinea split index for windy.

04:50This means rainy.

04:52So, we will calculate three different class labels of weather.

04:56Finally, weighted guinea for weather.

04:59Weighted guinea for weather is the probability of sunny

05:02into guinea value of sunny.

05:06Plus, probability of windy into guinea value of windy.

05:11Probability of rainy into guinea value of rainy.

05:14So, in this case, we will separate.

05:16So, sunny in the first place, there are three data.

05:19So, it is 3 by 10.

05:20That is the guinea value we provide.

05:230.44.

05:25So, we will multiply.

05:27Then, probability of windy data.

05:29Windy data, there are four windy data.

05:32So, 4 by 10.

05:33That is the guinea split index.

05:35So, in this case, we will provide.

05:36We have 41.4% of the disorders.

05:41So, we will check.

05:42This is the parent node.

05:44So, in the next parent node,

05:46we will choose the parent node.

05:47What classes are yes or no?

05:49So, first, yes.

05:50There are five instances.

05:51In five instances, we will decide.

05:53So, one minus five by five the whole square.

05:58So, one minus one is zero.

06:00So, this is the spitting proper.

06:02That is the guinea split index value is zero.

06:04That is the guinea split index of parents.

06:06So, the guinea split index of parents.

06:08Probability of yes into guinea split index of yes.

06:11Plus probability of no into guinea value of no.

06:13Check.

06:14Then, similarly, this is the money column.

06:18So, the money column is rich or poor.

06:20So, rich or poor based on decision column.

06:23Based on the guinea value.

06:24Check.

06:25This is the final decision of root node.

06:29So, the guinea split index of whether, parents and money.

06:33So, this is the minimum value of parents.

06:36So, the parents is the root node.

06:38Now, finally, I show the decision tree.

06:41So, this is the first layer of data.

06:45So, this is the first layer of parents.

06:48Parents are the root node.

06:50Yes.

06:51So, the guinea split index is zero.

06:54So, the guinea split index is zero.

06:55So, this is the first layer of data.

06:58You will see the other two written,

07:20Second, further spitting is nonno data.

07:24parents equal to no data

07:25parents equal to no data

07:27first parents base

07:29create

07:29next

07:30weather or money

07:32base

07:32create

07:33so

07:34we can consider

07:37so

07:38weather attribute

07:39parents equal to no data

07:40parents equal to no data

07:41weather attribute

07:43so

07:44weather attribute

07:45sunny

07:46rainy

07:46windy

07:47so

07:47separate

07:48two instance

07:49two instance

07:50one instance

07:51so

07:51sunny

07:51calculate

07:52rainy

07:53and windy

07:54separate

07:54guinea

07:54split

07:55index

07:55calculate

07:55sunny

07:56data

07:57consider

07:57two

07:58cinema

07:59tennis

07:59decide

08:00so

08:01guinea

08:02split

08:02value

08:02zero

08:03rainy

08:03consider

08:04stay

08:04in

08:05above

08:05value

08:050

08:06windy

08:06one

08:06cinema

08:08one

08:08shopping

08:09decide

08:10so

08:10guinea

08:11split

08:11index

08:12value

08:120.5

08:12this

08:13normal

08:14weighted

08:16guinea

08:16index

08:16calculate

08:17when

08:18guinea

08:18index

08:19of

08:19sunny

08:20into

08:21probability

08:21of sunny

08:22into

08:23guinea

08:24value of sunny

08:24provide

08:25and here

08:25is

08:25probability of sunny

08:27given

08:28parents

08:30equal to

08:31no

08:31so

08:32in the data

08:32we have

08:33probability value

08:35into

08:35guinea

08:38value of

08:38sunny

08:40given

08:42parents

08:42equal to

08:43no

08:43okay

08:44in the

08:44math

08:44plus

08:45probability of

08:47windy

08:48given

08:49parents

08:49equal to

08:50no

08:50we have

08:51separated

08:51calculate

08:52okay

08:53then

08:54check

08:55money

08:56column

08:57check

08:57once

08:57money

08:58column

08:58check

08:59and

09:00analyze

09:01we have

09:02We will analyze how the minimum value is.

09:04We will base weather based.

09:06In the next depth layer tool,

09:08we will create weather based.

09:10We will add no.

09:12We will add sunny, rainy and windy.

09:14We will create sunny based.

09:16We will decide both tennis based.

09:18We will decide.

09:20Guinea index value is 0.

09:22No further splitting.

09:24We will consider rainy data.

09:26We will add no further splitting.

09:28Guinea value index is 0.

09:30We will decide one of the cinema.

09:32We will decide shopping.

09:34Now, we will split the other.

09:36We will split the other.

09:38We will split the other.

09:40Money based.

09:42Who are rich.

09:44So, we will create a decision tree.

09:46Guinea split index based.

09:52This is the splitting criteria.

09:54We will see the next video.

Python - Splitting Criteria: Gini | Python Courses in Tamil | Skillfloor

Category

Transcript

Be the first to comment

Recommended